From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Kok, Auke" Subject: Re: [PATCH] pcnet: fix sparse triviality Date: Mon, 29 Oct 2007 10:54:25 -0700 Message-ID: <47261E51.2000301@intel.com> References: <20071026221132.1913.24518.stgit@localhost.localdomain> <4725B045.80705@garzik.org> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Cc: pcnet32@verizon.net, netdev@vger.kernel.org To: Jeff Garzik Return-path: Received: from mga03.intel.com ([143.182.124.21]:37244 "EHLO mga03.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752608AbXJ2SDd (ORCPT ); Mon, 29 Oct 2007 14:03:33 -0400 In-Reply-To: <4725B045.80705@garzik.org> Sender: netdev-owner@vger.kernel.org List-Id: netdev.vger.kernel.org Jeff Garzik wrote: > Auke Kok wrote: >> Since data can never exceed u32, it can't even be larger than >> LONG_MAX/HZ. >> >> Signed-off-by: Auke Kok >> Cc: pcnet32@verizon.net >> --- >> > Two comments: > > 1) I would prefer to pick a sane limit, like "1 day". The unit of > 'data' is seconds, so IMO we should not allow stupid timeouts, much less > INT_MAX ones :) But hey, then again, maybe we should permit root to > hang themselves with own rope... well, in this case it's only used for the blink interval for the LED on the back of the adapter. This is completely interruptable by the user with ^C when running ethtool, so it's not going to hang anything (traffic should continue nicely). yes INT_MAX is insanely long. moving all of this checking to the ethtool generic code makes much more sense indeed. > 2) [tangent] someone really should add the obvious ssleep_interruptible() *pass* :) Auke